[R] Functions in formulae ??
Jeremy Clark
jeremyclarkmel at gmail.com
Wed Nov 27 01:44:05 CET 2013
Dear All,
In the following simple case I can't seem to get an improved fit,
despite trying all of the control possibilities. As there seem to be
no examples anywhere which show use of functions such as "dnorm"
within a formula, and as I am not confident at all that my formula is
correctly configured, I would appreciate some feedback ! Many thanks
in advance.
library(minpack.lm)
gg<-c(170, 171, 172, 173, 174, 175,
176, 177, 178, 179, 180, 181, 182,
183, 184, 185, 186, 187, 188, 189,
190, 191, 192, 193, 194, 195, 196,
197, 198, 199, 200, 201, 202,
203, 204, 205, 206, 207, 208, 209)
dd<-c(1673, 1659, 1738, 1687, 1882, 2010, 2202,
2248, 2409, 2417, 2215, 2279, 2539, 2479, 2341,
2395, 2314, 2404, 2369, 2254, 2048, 1899,
1892, 1744, 1333, 1183, 982, 772, 526,
451, 335, 253, 157, 108, 67, 44,
26, 12, 10, 0)
dd100000 <- dd * 0.00001
dd100000gg <- data.frame(dd100000, gg)
## modLM4:
modLM4 <- nlsLM(dd100000 ~ dnorm(gg, mu,sigma, log = FALSE), data =
dd100000gg, start = c(mu = 190, sigma = 10), trace = TRUE)
windows()
## plot data
plot(gg, dd100000, main = "modLM4")
## plot fitted values
lines(gg, fitted(modLM4), col = 2, lwd = 2)
On Thu, Nov 21, 2013 at 11:57 AM, Jeremy Clark <jeremyclarkmel at gmail.com> wrote:
> Dear All,
>
> In the following simple case I can't seem to get an improved fit,
> despite trying all of the control possibilities. As there seem to be
> no examples anywhere which show use of functions such as "dnorm"
> within a formula, and as I am not confident at all that my formula is
> correctly configured, I would appreciate some feedback ! Many thanks
> in advance.
>
> library(minpack.lm)
>
> gg<-c(170, 171, 172, 173, 174, 175,
> 176, 177, 178, 179, 180, 181, 182,
> 183, 184, 185, 186, 187, 188, 189,
> 190, 191, 192, 193, 194, 195, 196,
> 197, 198, 199, 200, 201, 202,
> 203, 204, 205, 206, 207, 208, 209)
>
> dd<-c(1673, 1659, 1738, 1687, 1882, 2010, 2202,
> 2248, 2409, 2417, 2215, 2279, 2539, 2479, 2341,
> 2395, 2314, 2404, 2369, 2254, 2048, 1899,
> 1892, 1744, 1333, 1183, 982, 772, 526,
> 451, 335, 253, 157, 108, 67, 44,
> 26, 12, 10, 0)
>
> dd100000 <- dd * 0.00001
>
> dd100000gg <- data.frame(dd100000, gg)
>
> ## modLM4:
>
> modLM4 <- nlsLM(dd100000 ~ dnorm(gg, mu,sigma, log = FALSE), data =
> dd100000gg, start = c(mu = 190, sigma = 10), trace = TRUE)
>
> windows()
> ## plot data
> plot(gg, dd100000, main = "modLM4")
> ## plot fitted values
> lines(gg, fitted(modLM4), col = 2, lwd = 2)
--
MEDIMEL,
www.plastic-surgery-poland.co.uk
www.chirurgia-szczecin.pl
Jeremy Clark. Address: ul. Klonowica 45/2, 71-249 Szczecin, Poland.
More information about the R-help
mailing list